* [FFmpeg-devel] [PATCH] apv_decode: Replace division with shift
@ 2025-04-27 18:49 Mark Thompson
2025-04-27 22:21 ` James Almer
0 siblings, 1 reply; 3+ messages in thread
From: Mark Thompson @ 2025-04-27 18:49 UTC (permalink / raw)
To: ffmpeg-devel
The compiler can't see that this should be a shift and generates a real
division which is slow enough to appear in profiles on its own.
---
libavcodec/apv_decode.c | 12 ++++++------
1 file changed, 6 insertions(+), 6 deletions(-)
diff --git a/libavcodec/apv_decode.c b/libavcodec/apv_decode.c
index e28bc29c8f..e2c3161978 100644
--- a/libavcodec/apv_decode.c
+++ b/libavcodec/apv_decode.c
@@ -173,8 +173,8 @@ static int apv_decode_tile_component(AVCodecContext *avctx, void *data,
const AVPixFmtDescriptor *pix_fmt_desc =
av_pix_fmt_desc_get(avctx->pix_fmt);
- int sub_w = comp_index == 0 ? 1 : pix_fmt_desc->log2_chroma_w + 1;
- int sub_h = comp_index == 0 ? 1 : pix_fmt_desc->log2_chroma_h + 1;
+ int sub_w_shift = comp_index == 0 ? 0 : pix_fmt_desc->log2_chroma_w;
+ int sub_h_shift = comp_index == 0 ? 0 : pix_fmt_desc->log2_chroma_h;
APVRawTile *tile = &input->tile[tile_index];
@@ -190,8 +190,8 @@ static int apv_decode_tile_component(AVCodecContext *avctx, void *data,
int tile_mb_width = tile_width / APV_MB_WIDTH;
int tile_mb_height = tile_height / APV_MB_HEIGHT;
- int blk_mb_width = 2 / sub_w;
- int blk_mb_height = 2 / sub_h;
+ int blk_mb_width = 2 >> sub_w_shift;
+ int blk_mb_height = 2 >> sub_h_shift;
int bit_depth;
int qp_shift;
@@ -234,10 +234,10 @@ static int apv_decode_tile_component(AVCodecContext *avctx, void *data,
for (int blk_x = 0; blk_x < blk_mb_width; blk_x++) {
int frame_y = (tile_start_y +
APV_MB_HEIGHT * mb_y +
- APV_TR_SIZE * blk_y) / sub_h;
+ APV_TR_SIZE * blk_y) >> sub_h_shift;
int frame_x = (tile_start_x +
APV_MB_WIDTH * mb_x +
- APV_TR_SIZE * blk_x) / sub_w;
+ APV_TR_SIZE * blk_x) >> sub_w_shift;
ptrdiff_t frame_pitch = apv->output_frame->linesize[comp_index];
uint8_t *block_start = apv->output_frame->data[comp_index] +
--
2.47.2
_______________________________________________
ffmpeg-devel mailing list
ffmpeg-devel@ffmpeg.org
https://ffmpeg.org/mailman/listinfo/ffmpeg-devel
To unsubscribe, visit link above, or email
ffmpeg-devel-request@ffmpeg.org with subject "unsubscribe".
^ permalink raw reply [flat|nested] 3+ messages in thread* Re: [FFmpeg-devel] [PATCH] apv_decode: Replace division with shift
2025-04-27 18:49 [FFmpeg-devel] [PATCH] apv_decode: Replace division with shift Mark Thompson
@ 2025-04-27 22:21 ` James Almer
2025-04-30 22:09 ` Mark Thompson
0 siblings, 1 reply; 3+ messages in thread
From: James Almer @ 2025-04-27 22:21 UTC (permalink / raw)
To: ffmpeg-devel
[-- Attachment #1.1.1: Type: text/plain, Size: 2370 bytes --]
On 4/27/2025 3:49 PM, Mark Thompson wrote:
> The compiler can't see that this should be a shift and generates a real
> division which is slow enough to appear in profiles on its own.
> ---
> libavcodec/apv_decode.c | 12 ++++++------
> 1 file changed, 6 insertions(+), 6 deletions(-)
>
> diff --git a/libavcodec/apv_decode.c b/libavcodec/apv_decode.c
> index e28bc29c8f..e2c3161978 100644
> --- a/libavcodec/apv_decode.c
> +++ b/libavcodec/apv_decode.c
> @@ -173,8 +173,8 @@ static int apv_decode_tile_component(AVCodecContext *avctx, void *data,
> const AVPixFmtDescriptor *pix_fmt_desc =
> av_pix_fmt_desc_get(avctx->pix_fmt);
>
> - int sub_w = comp_index == 0 ? 1 : pix_fmt_desc->log2_chroma_w + 1;
> - int sub_h = comp_index == 0 ? 1 : pix_fmt_desc->log2_chroma_h + 1;
> + int sub_w_shift = comp_index == 0 ? 0 : pix_fmt_desc->log2_chroma_w;
> + int sub_h_shift = comp_index == 0 ? 0 : pix_fmt_desc->log2_chroma_h;
>
> APVRawTile *tile = &input->tile[tile_index];
>
> @@ -190,8 +190,8 @@ static int apv_decode_tile_component(AVCodecContext *avctx, void *data,
> int tile_mb_width = tile_width / APV_MB_WIDTH;
> int tile_mb_height = tile_height / APV_MB_HEIGHT;
>
> - int blk_mb_width = 2 / sub_w;
> - int blk_mb_height = 2 / sub_h;
> + int blk_mb_width = 2 >> sub_w_shift;
> + int blk_mb_height = 2 >> sub_h_shift;
>
> int bit_depth;
> int qp_shift;
> @@ -234,10 +234,10 @@ static int apv_decode_tile_component(AVCodecContext *avctx, void *data,
> for (int blk_x = 0; blk_x < blk_mb_width; blk_x++) {
> int frame_y = (tile_start_y +
> APV_MB_HEIGHT * mb_y +
> - APV_TR_SIZE * blk_y) / sub_h;
> + APV_TR_SIZE * blk_y) >> sub_h_shift;
> int frame_x = (tile_start_x +
> APV_MB_WIDTH * mb_x +
> - APV_TR_SIZE * blk_x) / sub_w;
> + APV_TR_SIZE * blk_x) >> sub_w_shift;
>
> ptrdiff_t frame_pitch = apv->output_frame->linesize[comp_index];
> uint8_t *block_start = apv->output_frame->data[comp_index] +
LGTM.
[-- Attachment #1.2: OpenPGP digital signature --]
[-- Type: application/pgp-signature, Size: 495 bytes --]
[-- Attachment #2: Type: text/plain, Size: 251 bytes --]
_______________________________________________
ffmpeg-devel mailing list
ffmpeg-devel@ffmpeg.org
https://ffmpeg.org/mailman/listinfo/ffmpeg-devel
To unsubscribe, visit link above, or email
ffmpeg-devel-request@ffmpeg.org with subject "unsubscribe".
^ permalink raw reply [flat|nested] 3+ messages in thread* Re: [FFmpeg-devel] [PATCH] apv_decode: Replace division with shift
2025-04-27 22:21 ` James Almer
@ 2025-04-30 22:09 ` Mark Thompson
0 siblings, 0 replies; 3+ messages in thread
From: Mark Thompson @ 2025-04-30 22:09 UTC (permalink / raw)
To: ffmpeg-devel
On 27/04/2025 23:21, James Almer wrote:
> On 4/27/2025 3:49 PM, Mark Thompson wrote:
>> The compiler can't see that this should be a shift and generates a real
>> division which is slow enough to appear in profiles on its own.
>> ---
>> libavcodec/apv_decode.c | 12 ++++++------
>> 1 file changed, 6 insertions(+), 6 deletions(-)
>>
>> diff --git a/libavcodec/apv_decode.c b/libavcodec/apv_decode.c
>> index e28bc29c8f..e2c3161978 100644
>> --- a/libavcodec/apv_decode.c
>> +++ b/libavcodec/apv_decode.c
>> @@ -173,8 +173,8 @@ static int apv_decode_tile_component(AVCodecContext *avctx, void *data,
>> const AVPixFmtDescriptor *pix_fmt_desc =
>> av_pix_fmt_desc_get(avctx->pix_fmt);
>> - int sub_w = comp_index == 0 ? 1 : pix_fmt_desc->log2_chroma_w + 1;
>> - int sub_h = comp_index == 0 ? 1 : pix_fmt_desc->log2_chroma_h + 1;
>> + int sub_w_shift = comp_index == 0 ? 0 : pix_fmt_desc->log2_chroma_w;
>> + int sub_h_shift = comp_index == 0 ? 0 : pix_fmt_desc->log2_chroma_h;
>> APVRawTile *tile = &input->tile[tile_index];
>> @@ -190,8 +190,8 @@ static int apv_decode_tile_component(AVCodecContext *avctx, void *data,
>> int tile_mb_width = tile_width / APV_MB_WIDTH;
>> int tile_mb_height = tile_height / APV_MB_HEIGHT;
>> - int blk_mb_width = 2 / sub_w;
>> - int blk_mb_height = 2 / sub_h;
>> + int blk_mb_width = 2 >> sub_w_shift;
>> + int blk_mb_height = 2 >> sub_h_shift;
>> int bit_depth;
>> int qp_shift;
>> @@ -234,10 +234,10 @@ static int apv_decode_tile_component(AVCodecContext *avctx, void *data,
>> for (int blk_x = 0; blk_x < blk_mb_width; blk_x++) {
>> int frame_y = (tile_start_y +
>> APV_MB_HEIGHT * mb_y +
>> - APV_TR_SIZE * blk_y) / sub_h;
>> + APV_TR_SIZE * blk_y) >> sub_h_shift;
>> int frame_x = (tile_start_x +
>> APV_MB_WIDTH * mb_x +
>> - APV_TR_SIZE * blk_x) / sub_w;
>> + APV_TR_SIZE * blk_x) >> sub_w_shift;
>> ptrdiff_t frame_pitch = apv->output_frame->linesize[comp_index];
>> uint8_t *block_start = apv->output_frame->data[comp_index] +
>
> LGTM.
Applied.
Thanks,
- Mark
_______________________________________________
ffmpeg-devel mailing list
ffmpeg-devel@ffmpeg.org
https://ffmpeg.org/mailman/listinfo/ffmpeg-devel
To unsubscribe, visit link above, or email
ffmpeg-devel-request@ffmpeg.org with subject "unsubscribe".
^ permalink raw reply [flat|nested] 3+ messages in thread
end of thread, other threads:[~2025-04-30 22:09 UTC | newest]
Thread overview: 3+ messages (download: mbox.gz / follow: Atom feed)
-- links below jump to the message on this page --
2025-04-27 18:49 [FFmpeg-devel] [PATCH] apv_decode: Replace division with shift Mark Thompson
2025-04-27 22:21 ` James Almer
2025-04-30 22:09 ` Mark Thompson
Git Inbox Mirror of the ffmpeg-devel mailing list - see https://ffmpeg.org/mailman/listinfo/ffmpeg-devel
This inbox may be cloned and mirrored by anyone:
git clone --mirror https://master.gitmailbox.com/ffmpegdev/0 ffmpegdev/git/0.git
# If you have public-inbox 1.1+ installed, you may
# initialize and index your mirror using the following commands:
public-inbox-init -V2 ffmpegdev ffmpegdev/ https://master.gitmailbox.com/ffmpegdev \
ffmpegdev@gitmailbox.com
public-inbox-index ffmpegdev
Example config snippet for mirrors.
AGPL code for this site: git clone https://public-inbox.org/public-inbox.git